Stereochemical control in microbial reduction. Part 31: Reduction of alkyl 2-oxo-4-arylbutyrates by baker's yeast under selected reaction conditions

Article abstract of DOI:10.1016/S0957-4166(98)00277-8

Treatment of baker's yeast with phenacyl chloride in an aqueous-organic solvent has been proven to be an effective method of inhibiting the enzymes that afford (S)-enantiomers of α-hydroxy esters in the reduction of α-keto esters. The procedure is effective for the whole-cell system to produce the (R)-product with high chemical yield and high enantiomeric excess.

Full text of DOI:10.1016/S0957-4166(98)00277-8

TETRAHEDRON:
ASYMMETRY
Pergamon
Tetrahedron: Asymmetry 9 (1998) 2725–2737
Stereochemical control in microbial reduction. Part 31: Reduction
of alkyl 2-oxo-4-arylbutyrates by baker’s yeast under selected
reaction conditions
Duc Hai Dao,a Mutsuo Okamura,a Takeshi Akasaka,a Yasushi Kawai,b Kouichi Hida b and
Atsuyoshi Ohno b,∗
aCourse of Fundamental Studies and Energy Technology, Graduate School of Science and Technology, Niigata University,
Ikarashi, Niigata 950-2181, Japan
bInstitute for Chemical Research, Kyoto University, Uji, Kyoto 611-0011, Japan
Received 22 June 1998; accepted 8 July 1998
Abstract
Treatment of baker’s yeast with phenacyl chloride in an aqueous–organic solvent has been proven to be an
effective method of inhibiting the enzymes that afford (S)-enantiomers of α-hydroxy esters in the reduction of
α-keto esters. The procedure is effective for the whole-cell system to produce the (R)-product with high chemical
yield and high enantiomeric excess. © 1998 Elsevier Science Ltd. All rights reserved.
1. Introduction
The asymmetric production of bioactive compounds starting from chiral building blocks has been
well documented for their advantages.1–3 The availability and cost of these chiral reagents is highly
dependent on their method of production, and much effort has been made in the common interest of
obtaining cheaper optically active compounds of high purity and yield.1,4 One prime example is the
preparation of the chiral reagent alkyl (R)-2-hydroxy-4-phenylbutanoate, a versatile key intermediate for
the synthesis of a variety of angiotensin converting enzyme (ACE) inhibitors from their corresponding
prochiral precursors, alkyl 2-oxo-4-phenylbutanoates.5–7 Often, the biotechnological preparation of
chiral molecules such as alkyl (R)-2-hydroxy-4-phenylbutanoates5,8–10 is the method of choice rather
than just the chemical approach.11,12 Microbial reduction of ethyl 2-oxo-4-phenylbutanoate to ethyl (R)-
2-hydroxy-4-phenylbutanoate with excellent chemical yield (90%) and ee (>99%) under the catalysis of
Daucus carota cells has already been reported.10 The same reaction under the catalysis of baker’s yeast
∗
Corresponding author. Fax: +81-774-38-3209; e-mail: ohno@boc.kuicr.kyoto-u.ac.jp
0957-4166/98/$ - see front matter © 1998 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(98)00277-8

2726
D. H. Dao et al. / Tetrahedron: Asymmetry 9 (1998) 2725–2737
(Saccharomyces cerevisiae), which has long been recognized as a valuable reagent in organic asymmetric
synthesis, has also been reported recently by us to give high chemical yield (80–90%) with excellent ee
(>90%)13 by selecting reaction conditions such as medium, temperature, addition of a third reagent and
so on.14–17
In order to extend the scope of the reduction, we studied the reduction of alkyl 2-oxo-4-arylbutanoates
1 to the corresponding alkyl 2-hydroxy-4-arylbutanoates 2 mediated by baker’s yeast (Scheme 1) under
the influence of the third reagent added to the reaction system.13 We now wish to report the effect of
substituents on the phenyl ring with respect to chemical yield and ee of the product. In addition, we
would also like to exploit the potential of our technique in the yeast-mediated reduction of other aromatic
α-keto esters such as alkyl phenylformates and phenylpyruvates. It has been reported that, although the
reduction of phenylformate affords (R)-mandelate in 60–70% chemical yield with 98% ee, the reduction
of phenylpyruvate affords racemic product in 40–50% chemical yield.18 Of course, the disadvantages in
chemical yield and enantioselectivity have been resolved with the use of isolated enzyme systems.19,20
Scheme 1.
2. Results and discussion
2.1. Syntheses of materials
Ethyl 2-oxo-4-phenylbutanoate substituted by o-, m- and p-methyl, o-, m- and p-chloro, and p-
fluoro, as well as p-methoxy groups were all prepared by converting the alcohols to the corresponding

D. H. Dao et al. / Tetrahedron: Asymmetry 9 (1998) 2725–2737
2727
bromides,21–24 which were then coupled with diethyl oxalate through the Grignard reactions.25,26 Propyl
2-oxo-4-p-fluorophenylbutanoate 1g and isobutyl 2-oxo-4-p-fluorophenylbutanoate 1h were prepared
by acid-catalyzed exchange reactions from ethyl 2-oxo-4-p-fluorophenylbutanoate 1f with appropriate
alcohols, respectively. Ethyl 2-oxo-4-p-nitrophenylbutanoate 1m was prepared by oxidation of ethyl 2-
hydroxy-4-p-nitrophenylbutanoate 2m with pyridinium dichromate,27,28 which was prepared from 4-(p-
nitrophenyl)butyric acid by the Hell–Volhard–Zelinski procedure.29,30
2.2. Determination of configuration
The configuration of the o-, m- and p-chloro compounds 2i, j and k were determined by reductive
dechlorination at the aromatic moiety using Pd/C.31,32 The configurations of the products 2a were
elucidated by comparing their retention times on HPLC with a chiral stationary phase to that of the
authentic sample of the (R)-configuration. The m-methyl, p-methoxy, p-fluoro and p-nitro derivatives 2c,
e, f and m were initially acetylated to protect the α-hydroxy group and oxidized by ruthenium tetroxide
to give the monoethyl α-acetoxy glutarate.33–36 The monoethyl ester was converted to the corresponding
p-nitrobenzoate 4, through diethyl α-hydroxy glutarate 3, for convenience in chromatography. The
configuration of 4 thus obtained was then compared to that derived from 2a.
Liberation of the secondary alcohol moiety by removal of the p-nitrobenzoyl group or standing 3 in
pyridine inherently causes intramolecular cyclization to yield a lactone 5, another useful chiral building
block, in 76–90% chemical yield (Scheme 2). The chiral lactone 5 is also available by yeast-mediated
reduction of 3-benzoylpropionic acid37,38 followed by the oxidation with ruthenium tetroxide.33–36
Scheme 2.

2728
D. H. Dao et al. / Tetrahedron: Asymmetry 9 (1998) 2725–2737
2.3. Substituent effect
Substitution of an electron-withdrawing group retards the reaction as seen in Table 1. It is interesting
to note that a chlorine substituent at any position of the phenyl ring exerts good stereoselectivity. The
effect of the steric size of a molecule is also seen in the reductions of propyl and isobutyl esters 1g and h
when they are compared to the reduction of the corresponding ethyl ester, 1f.13
When the microbe is incubated in the presence of phenacyl chloride, stereochemical results are
improved appreciably for all the substrates studied, as has been reported previously for the reduction
of unsubstituted substrate, 1a.13 It is obvious that phenacyl chloride inactivates the enzyme(s) that afford
the (S)-isomer of the product, because, in most cases, an increase in the stereoselectivity of the (R)-
isomer is improved at the sacrifice of chemical yield after prolonged incubation time with this reagent. In
addition, since we know that changing the ethyl group at the alcohol moiety to a butyl or a higher alkyl
group improves the selectivity to a satisfactory level (90% ee), we believe that the present results appear
satisfactory except for 1b, a substrate with an ortho-substituent. Clearly, the selectivity results, in the
presence of phenacyl chloride, are satisfactory for the m-methyl-, p-methoxy-, p-fluoro- and m- and p-
chloro-substituents, whilst the effect is less sensitive for the p-methyl- and p-nitro-substituents. It seems
that a moderately bulky group (or atom) at the para-position is required to be sensitive to the effects of
phenacyl chloride. Electronic effects from the substituents on the phenyl ring seem to play little part in
controlling the selectivity under the influence of phenacyl chloride. Although an electron-withdrawing
p-fluoro-substituted substrate with a bulky alcohol moiety, 1h, exerts excellent stereoselectivity, the
chemical yield from this substrate is far from being satisfactory. Because the chemical yield is defined
as that after 24 h of reaction and we observed no other product than 2 on HPLC, there is no doubt that
prolonged reaction time and/or an increase in the amount of baker’s yeast may improve the chemical
yield to a satisfactory level.
2.4. Effect of carbon chain length
The use of phenacyl chloride as a ‘stereochemical regulator’ has so far demonstrated its potential
effectiveness in controlling the stereochemical outcome of the reduction. To verify further the effect-
iveness of this reagent, we tested its ability to effect the stereochemical control of the baker’s yeast
reduction of other α-keto esters having different carbon chain length in the acid moiety. The results are
summarized in Table 2, which again proves that the use of a stereochemical regulator is an effective
method for controling the stereochemical result of the reduction.
Although methyl and ethyl benzoylformates are reduced with high ee under standard conditions,
addition of phenacyl chloride has greatly enhanced the selectivity further. The length of the carbon chain
in the acid moiety of the esters so far studied does not affect the stereochemical result or chemical
yield appreciably: they are all satisfactory. Thus, it has been proven that the treatment of the yeast
cells in aqueous diethyl ether in the presence of phenacyl chloride is an effective device to control
the stereochemical course of the reduction, affording the corresponding (R)-enantiomer of the hydroxy
ester. Since the reactivity of the microbe decreases, there is no doubt that phenacyl chloride inhibits the
enzyme(s) responsible for the production of the (S)-enantiomer of the ester.

D. H. Dao et al. / Tetrahedron: Asymmetry 9 (1998) 2725–2737
2729
Table 1
Effect of substituent on stereoselectivity in the reduction of alkyl 2-oxo-4-arylbutanoate 1 mediated
by baker’s yeast

2730
D. H. Dao et al. / Tetrahedron: Asymmetry 9 (1998) 2725–2737
Table 2
Reduction of α-keto esters having different carbon chain lengths
Table 3
Medium-scale preparation of α-hydroxy estersa
2.5. Medium-scale preparation
It is important to confirm that the method is practically applicable to organic syntheses, and the
reduction was run with a substrate of laboratory-scale amount. Results, summarized in Table 3, have
been found satisfactory.
3. Experimental section
3.1. Instruments
Liquid-phase chromatograms (HPLC) were obtained on a Hitachi 655 using a Chiralcel OD 25+5
(Daicel) with hexane:isopropanol as the eluent. 1H, 13C and 19F NMR spectra were recorded on a Varian
VXR 200 FT-NMR spectrometer in CDCl3 (TMS and Cl3CF as internal standards). Optical rotations

D. H. Dao et al. / Tetrahedron: Asymmetry 9 (1998) 2725–2737
2731
were measured on a JASCO DIP-181 digital polarimeter. Infrared analysis was performed with a JASCO
FT-IR 5300 spectrophotometer. Elemental analyses were conducted using a Yanaco MT-5 elemental
analyzer. Melting points were measured with a Yanagimoto micro melting point apparatus, and were
not corrected.
3.2. Materials and methods
Reagents were purchased from Nacalai Tesque Inc., Wako Pure Chemical Industries Ltd, and the
Aldrich Chemical Co. unless otherwise stated. Baker’s yeast (Saccharomyces cerevisiae) was purchased
from the Oriental Yeast Co. and stored in a refrigerator. Column chromatography was performed on
silica gel using the solvent mixtures indicated. Both diethyl ether and THF were dried over CaCl2 and
finally distilled over Na with benzophenone ketyl as an indicator. Benzene and CH2Cl2 were distilled on
CaH2 or dried over 4 Å molecular sieves. All other solvents were of analytical grade. Dried and distilled
solvents were employed whenever necessary. Preparation of 2-oxo-4-phenylbutanoate, 1a, was described
in a previous paper from our laboratory.13 The purity of all the reagents employed for the reactions were
confirmed by 1H NMR spectroscopy and gave satisfactory results.
3.3. Preparation alkyl 2-oxo-4-arylbutanoates
As a typical run, (4-fluorophenyl)acetic acid was reduced to 2-(4-fluorophenyl)ethyl alcohol by
LiAlH4, and the alcohol was converted into the corresponding bromide by bromination with PBr3.21–24
2-(4-Fluorophenyl)ethyl alcohol: 1H NMR (δ from TMS in CDCl3): 2.20 (bs, 1H), 2.84 (t, 2H, J=6.6
Hz), 3.84 (t, 2H, J=6.6 Hz), 6.99 (m, 2H), 7.18 (m, 2H).
1
2-(4-Fluorophenyl)ethyl bromide: H NMR (δ from TMS in CDCl3): 2.95 (t, 2H, J=7.5 Hz), 3.39 (t,
2H, J=7.5 Hz), 7.01 (m, 2H), 7.18 (m, 2H).
To a 100 ml two-necked round-bottomed flask containing 0.38 g (15.6 mmol) Mg equipped with a
magnetic stirrer and a Dimroth condenser connected to an argon balloon was added a solution of 3.4
g (16.7 mmol) of the halide in 20 ml diethyl ether via a syringe. External heating with good agitation
may be needed to initiate the reaction. The solution was then refluxed for 30–40 min with gentle heating;
it was then cooled to room temperature before conveying it to a second vessel, containing 1.9 g (13.0
mmol) of diethyl oxalate in 20 ml diethyl ether, equipped with a magnetic stirrer, a dropping funnel and a
Dimroth condenser protected from the air by a CaCl2 tube. The Grignard reagent was discharged slowly
into the reaction mixture with vigorous stirring in an ice-bath (1 h), then for a further 10 h or overnight
at room temperature, and was completed with 30–40 min reflux. Ethyl 2-oxo-(4-fluorophenyl)butanoate,
1f, (78% yield) was liberated after acidification of the reaction mixture with 2 M (1 M=1 mol dm−3) HCl
at 0°C and usual work-up.25,26 The ester was purified on a silica gel column with hexane:EtOAc=5:1 as
1
an eluent (Rf=0.51). H NMR (δ from TMS in CDCl3): 1.35 (t, 3H, J=7.2 Hz), 2.94 (t, 2H, J=7.1 Hz),
3.15 (t, 2H, J=7.1 Hz), 4.34 (q, 2H, J=7.2 Hz), 6.95 (m, 2H), 7.15 (m, 2H). Anal. calcd for C12H13FO3:
C, 64.28; H, 5.84%. Found: C, 64.37; H, 5.88%.
The propyl and isobutyl 2-oxo-(4-fluorophenyl)butanoates, 1g and 1h, were prepared by ester ex-
change reactions from the ethyl ester and the corresponding alcohols, respectively.
1
1g: H NMR (δ from TMS in CDCl3): 0.97 (t, 3H, J=7.2 Hz), 1.75 (m, 2H), 2.96 (t, 2H, J=7.2 Hz),
3.15 (t, 2H, J=6.8 Hz), 4.20 (t, 2H, J=6.8 Hz), 6.95 (m, 2H) and 7.15 (m, 2H). Anal. calcd for C13H15FO3:
C, 65.54; H, 6.35%. Found: C, 65.64; H, 6.39%.

2732
D. H. Dao et al. / Tetrahedron: Asymmetry 9 (1998) 2725–2737
1h: 1H NMR (δ from TMS in CDCl3): 0.96 (d, 6H, J=7.0 Hz), 2.03 (m, 1H), 2.93 (t, 2H, J=7.1 Hz),
3.15 (t, 2H, J=7.1 Hz), 4.01 (d, 2H, J=7.1 Hz), 6.96 (m, 2H) and 7.15 (m, 2H). Anal. calcd for C14H17FO3:
C, 66.65; H, 6.79%. Found: C, 66.48; H, 6.91%.
All other esters were prepared by methods similar to above in 30–78% yields.
1
1b: H NMR (δ from TMS in CDCl3): 1.35 (t, 3H, J=7.1 Hz), 2.32 (s, 3H), 2.94 (m, 2H), 3.13 (m,
2H), 4.31 (q, 2H, J=7.1 Hz) and 7.14 (s, 4H). Anal. calcd for C13H16O3: C, 70.89; H, 7.32%. Found: C,
70.93; H, 7.44%.
1
1c: H NMR (δ from TMS in CDCl3): 1.28 (t, 3H, J=7.1 Hz), 2.31 (s, 3H), 2.92 (t, 2H, J=7.1 Hz),
3.12 (t, 2H, 7.1 Hz), 4.20 (q, 2H, J=7.1 Hz), 6.99 (m, 3H) and 7.16 (m, 1H). Anal. calcd for C13H16O3:
C, 70.89; H, 7.32%. Found: C, 70.77; H, 7.38%.
1
1d: H NMR (δ from TMS in CDCl3): 1.34 (t, 3H, J=7.1 Hz), 2.31 (s, 3H), 2.91 (t, 2H, J=7.1 Hz),
3.15 (t, 2H, J=7.1 Hz), 4.29 (q, 2H, J=7.1 Hz) and 7.09 (m, 4H). Anal. calcd for C13H16O3: C, 70.89; H,
7.32%. Found: C, 71.02; H, 7.43%.
1e: 1H NMR (δ from TMS in CDCl3): 1.35 (t, 3H, J=7.1 Hz), 2.90 (t, 2H, J=7.2 Hz), 3.14 (t, 2H, J=7.1
Hz), 3.78 (s, 3H), 4.30 (q, 2H, J=7.2 Hz), 6.83 (d, 2H, J=8.6 Hz), 7.11 (d, 2H, J=8.6 Hz). Anal. calcd for
C13H16O4: C, 66.09; H, 6.83%. Found: C, 66.28; H, 6.96%.
1i: 1H NMR (δ from TMS in CDCl3): 1.35 (t, 3H, J=7.1 Hz), 3.05 (t, 2H, J=6.6 Hz), 3.17 (t, 2H, J=6.6
Hz), 4.31 (q, 2H, J=7.1 Hz) and 7.12–7.38 (m, 4H). Anal. calcd for C12H13ClO3: C, 59.88; H, 5.44%.
Found: C, 60.10; H, 5.39%.
1j: 1H NMR (δ from TMS in CDCl3): 1.35 (t, 3H, J=7.1 Hz), 2.93 (t, 2H, J=7.2 Hz), 3.18 (t, 2H, J=7.2
Hz), 4.31 (q, 2H, J=7.1 Hz), 7.06 (m, 1H) and 7.20 (m, 3H). Anal. calcd for C12H13ClO3: C, 59.88; H,
5.44%. Found: C, 60.02; H, 5.57%.
1
1k: H NMR (δ from TMS in CDCl3): 1.34 (t, 3H, J=7.2 Hz), 2.92 (t, 2H, J=7.2 Hz), 3.16 (t, 2H,
J=7.2 Hz), 4.32 (q, 2H, J=7.2 Hz), 7.12 (d, 2H) and 7.25 (d, 2H). Anal. calcd for C12H13ClO3: C, 59.88;
H, 5.44%. Found: C, 60.11; H, 5.53%.
1m: To a 100 ml round-bottomed flask equipped with a magnetic stirring bar was added 1.0 g (4.0
mmol) of ethyl 2-hydroxy-(4-nitrophenyl)butanoate 2m, 40 ml of CH2Cl2 and 9.0 g (24.0 mmol) of
pyridinium dichromate.27,28 The reaction, under vigorous stirring at room temperature, was run for 3.5
days to reach near completion as gauged by TLC (Rf=0.45, hexane:EtOAc=3:1) to afford, after work-up
and through silica gel column chromatography, ethyl 2-oxo-(4-nitrophenyl)butanoate 1m in 34% yield.
1H NMR (δ from TMS in CDCl3): 1.29 (t, 3H, J=7.1 Hz), 2.81 (t, 2H, J=6.5 Hz), 3.38 (t, 2H, J=6.5 Hz),
4.18 (q, 2H, J=7.1 Hz), 8.14 (d, 2H, J=8.9 Hz) and 8.31 (d, 2H, J=8.9 Hz). Anal. calcd for C12H13NO5:
C, 57.37; H, 5.22; N, 5.58%. Found: C, 57.39; H, 5.28; N, 5.52%.
Methyl and ethyl benzoylformate were prepared by an acid catalyzed exchange reaction of benzoyl-
formic acid (2.5 g, 16.7 mmol) with appropriate alcohols (15 ml) in near-quantitative yields after work-up
(silica gel column, hexane:EtOAc=5:1, Rf=0.50 and 0.57, respectively). 1H NMR spectra are identical to
the authentic samples.
Ethyl phenylpyruvate was obtained in 40% yield by the Grignard reaction25,26 with benzyl chloride
1
(5.0 g, 39.5 mmol) and diethyl oxalate (5.2 g, 35.6 mmol). H NMR (δ from TMS in CDCl3): 1.36 (t,
3H, J=7.2 Hz), 4.34 (q, 2H, J=7.2 Hz), 6.45 (s, 1H), 7.21–7.42 (m, 4H) and 7.78 (d, 2H, J=7.3 Hz).
Ethyl 2-oxo-5-phenylpentanoate in 67% yield was prepared from 1-bromo-3-phenylpropane (6.0 g,
1
30.1 mmol) and diethyl oxalate (3.7 g, 25.3 mmol) in a similar manner. H NMR (δ from TMS in
CDCl3): 1.29 (t, 3H, J=7.1 Hz), 1.92 (m, 2H), 2.59 (t, 2H, J=7.5 Hz), 2.81 (t, 2H, J=7.3 Hz), 4.25 (q, 2H,
J=7.1 Hz) and 7.15 (m, 5H). Anal. calcd for C13H16O3: C, 70.89; H, 7.32%. Found: C, 70.76; H, 7.40%.

D. H. Dao et al. / Tetrahedron: Asymmetry 9 (1998) 2725–2737
2733
3.4. Preparation of alkyl 2-hydroxy-4-arylbutanoates
As a typical run, ethyl 2-hydroxy-(4-fluorophenyl)butanoate, 2f, was prepared in 76% yield by NaBH4
reduction of 1f in THF at 0°C under stirring for 4–6 h. 1H NMR (δ from TMS in CDCl3): 1.26 (t, 3H,
J=7.2 Hz), 1.81–1.98 (m, 1H), 1.99–2.18 (m, 1H), 2.74 (t, 2H, J=7.2 Hz), 2.95 (bs, 1H), 4.15 (m, 1H),
4.21 (q, 2H, J=7.2 Hz), 6.96 (m, 2H), 7.15 (m, 2H): 19F NMR (δ from Cl3CF in CDCl3): −117.6. Anal.
calcd for C12H15FO3: C, 63.71; H, 6.68%. Found: C, 63.75; H, 6.60%.
1
2g: H NMR (δ from TMS in CDCl3): 0.95 (t, 3H, J=7.2 Hz), 1.66 (m, 2H), 1.81–1.98 (m, 1H),
1.99–2.17 (m, 1H), 2.74 (m, 2H), 2.84 (bs, 1H), 4.13 (t, 2H, J=7.2 Hz), 4.16 (m, 1H), 6.96 (m, 2H) and
7.14 (m, 2H). Anal. calcd for C13H17FO3: C, 64.99; H, 7.13%. Found: C, 65.04; H, 7.21%
2h: 1H NMR (δ from TMS in CDCl3): 0.95 (d, 6H, J=7.0 Hz), 1.83–1.98 (m, 2H), 1.99–2.18 (m, 1H),
2.74 (m, 2H), 2.87 (bs, 1H), 3.96 (d, 2H, J=7.0 Hz), 4.17 (m, 1H), 6.95 (m, 2H) and 7.15 (m, 2H). Anal.
calcd for C14H19FO3: C, 66.12; H, 7.53%. Found: C, 66.32; H, 7.58%.
All other phenyl substituted esters were prepared by methods similar to above in 56–85% yields.
2b: 1H NMR (δ from TMS in CDCl3): 1.29 (t, 3H, J=7.1 Hz), 1.81–1.98 (m, 1H), 1.99–2.16 (m, 1H),
2.32 (s, 3H), 2.75 (m, 2H), 2.91 (bs, 1H), 4.21 (m, 1H), 4.23 (q, 2H, J=7.1 Hz) and 7.13 (m, 4H). Anal.
calcd for C13H18O3: C, 70.25; H, 8.16%. Found: C, 70.34; H, 8.25%.
2c: 1H NMR (δ from TMS in CDCl3): 1.27 (t, 3H, J=7.1 Hz), 1.83–1.98 (m, 1H), 1.99–2.19 (m, 1H),
2.31 (s, 3H), 2.73 (t, 2H, J=7.2 Hz), 2.91 (bs, 1H), 4.15 (m, 1H), 4.20 (q, 2H, J=7.1 Hz), 6.98 (m, 3H)
and 7.15 (m, 1H). Anal. calcd for C13H18O3: C, 70.25; H, 8.16%. Found: C, 70.28; H, 8.18%.
2d: 1H NMR (δ from TMS in CDCl3): 1.29 (t, 3H, J=7.1 Hz), 1.81–1.98 (m, 1H), 1.99–2.23 (m, 1H),
2.31 (s, 3H), 2.74 (m, 2H), 2.83 (bs, 1H), 4.16 (m, 1H), 4.20 (q, 2H, J=7.1 Hz) and 7.10 (m, 4H). Anal.
calcd for C13H18O3: C, 70.25; H, 8.16%. Found: C, 70.43; H, 8.17%.
2e: 1H NMR (δ from TMS in CDCl3): 1.29 (t, 3H, J=7.1 Hz), 1.81–1.98 (m, 1H), 1.99–2.18 (m, 1H),
2.70 (t, 2H, J=7.4 Hz), 2.85 (bs, 1H), 3.78 (s, 3H), 4.16 (m, 1H), 4.21 (q, 2H, J=7.1 Hz), 6.83 (d, 2H,
J=8.6 Hz) and 7.11 (d, 2H, J=8.6 Hz). Anal. calcd for C13H18O4: C, 65.53; H, 7.61%. Found: C, 65.74;
H, 7.72%.
2i: 1H NMR (δ from TMS in CDCl3): 1.29 (t, 3H, J=7.1 Hz), 1.86–1.99 (m, 1H), 2.00–2.20 (m, 1H),
2.88 (t, 2H, J=7.9 Hz), 2.98 (bs, 1H), 4.20 (m, 1H), 4.21 (q, 2H, J=7.1 Hz) and 7.09–7.27 (m, 4H). Anal.
calcd for C12H15ClO3: C, 59.39; H, 6.23%. Found: C, 59.80; H, 6.27%.
2j: 1H NMR (δ from TMS in CDCl3): 1.30 (t, 3H, J=7.1 Hz), 1.83–1.99 (m, 1H), 2.00–2.19 (m, 1H),
2.74 (t, 2H, J=8.4 Hz), 2.96 (bs, 1H), 4.15 (m, 1H), 4.21 (q, 2H, J=7.1 Hz), 7.07 (m, 1H) and 7.19 (m,
3H). Anal. calcd for C12H15ClO3: C, 59.39; H, 6.23%. Found: C, 59.58; H, 6.25%.
2k: 1H NMR (δ from TMS in CDCl3): 1.28 (t, 3H, J=7.2 Hz), 1.81–1.98 (m, 1H), 1.99–2.17 (m, 1H),
2.73 (t, 2H, J=6.8 Hz), 2.89 (bs, 1H), 4.13 (m, 1H), 4.20 (q, 2H, J=7.2 Hz), 7.12 (m, 2H) and 7.26 (m,
2H). Anal. calcd for C12H15ClO3: C, 59.39; H, 6.23%. Found: C, 59.40; H, 6.27%.
2m: To a 200 ml round-bottomed flask fitted with a mechanical stirrer and a Dimroth condenser was
added 50 ml of benzene, 10 g (48.0 mmol) of 4-(4-nitrophenyl)butyric acid, 0.6 ml (0.8 g, 5.82 mmol) of
phosphorous trichloride and 2.5 ml (7.64 g, 48.0 mmol) of bromine. The resulting solution was refluxed
for 1–2 days until the initial bromine color was discharged. At this stage the benzene was distilled off
by replacing the condenser with a collecting side-arm. The removal of the solvent left a black oil which,
on cooling to room temperature, was added slowly to 3 M NaOH (150 ml) under good stirring in an ice-
bath. The reaction was run for 8–10 h at 0°C and then overnight at room temperature. The solution was
transferred to a 500 ml conical flask and was acidified with slow addition of 2 M HCl at 0°C. Extraction
with diethyl ether (5×50 ml) and usual work-up gave crude 2-hydroxy-4-(4-nitrophenyl)butyric acid

2734
D. H. Dao et al. / Tetrahedron: Asymmetry 9 (1998) 2725–2737
(Rf=0.10, eluent: hexane:EtOAc=1:2). 1H NMR (δ from TMS in CDCl3): 1.86–2.03 (m, 1H), 2.04–2.22
(m, 1H), 2.77–2.93 (m, 2H), 4.16 (bs, 1H), 7.35 (d, 2H, J=8.9 Hz) and 8.14 (d, 2H, J=8.9 Hz)
The product was difficult to purify by column chromatography on silica gel with several organic
solvent mixtures as eluents or by recrystallization, due to contamination caused by the remaining α-
bromo acid and 3-(4-nitrophenyl)propanol [1H NMR (δ from TMS in CDCl3): 2.02 (m, 2H), 2.40 (t, 2H,
J=6.8 Hz), 2.89 (t, 2H, J=7.0 Hz), 7.35 (d, 2H, J=8.9 Hz) and 8.15 (d, 2H, J=8.9 Hz), Rf=0.25] formed
by decarboxylation from 4-(4-nitrophenyl)butyric acid. Hence, acid catalyzed esterification of the crude
α-hydroxy acid in 30–40 ml anhydrous ethanol afforded 2m after usual work-up in 40% yield,29,30 which
was purified much more easily than the corresponding acid by column chromatography on silica gel with
1
hexane:EtOAc=3:1 as the eluent (Rf=0.29). H NMR (δ from TMS in CDCl3): 1.31 (t, 3H, J=7.1 Hz),
1.88–2.05 (m, 1H), 2.05–2.24 (m, 1H), 2.78–2.95 (m, 2H), 4.17 (m, 1H), 4.18–4.30 (q, 2H, J=7.1 Hz),
7.37 (d, 2H, J=8.9 Hz) and 8.15 (d, 2H, J=8.9 Hz); mp: 60–62°C. Anal. calcd for C12H15NO5: C, 56.91;
H, 5.97; N, 5.53%. Found: C, 56.88; H, 5.98; N, 5.60%.
The methyl and ethyl mandelates were obtained commercially. Similarly, D- and L-ethyl phenyl-
lactate were obtained from D- and L-phenyllactic acids, without racemization, in quantitative yields
(hexane:EtOAc=5:1, Rf=0.35). 1H NMR (δ from TMS in CDCl3): 1.26 (t, 3H, J=7.2 Hz), 2.78 (bs, 1H),
2.94 (dd, 1H, J=13.8 Hz, J=6.8 Hz), 3.13 (dd, 1H, J=13.8 Hz, J=4.8 Hz), 4.21 (q, 2H, J=7.2 Hz), 4.43
(dd, 1H, J=6.8 Hz, J=4.8 Hz) and 7.25 (m, 5H).
Ethyl 5-phenyl-2-hydroxypentanoate was prepared by the reduction of the corresponding keto ester
with NaBH4, as mentioned above, in 82% yield after the work-up [silica gel column, hexane:EtOAc=5:1,
1
3:1 then 1:1, Rf (5:1)=0.40]. H NMR (δ from TMS in CDCl3): 1.21 (t, 3H, J=7.1 Hz), 1.60–1.88 (m,
4H), 2.58 (m, 2H), 2.74 (bs, 1H), 4.17 (m, 1H), 4.19 (q, 2H, J=7.1 Hz) and 7.18 (m, 5H). Anal. calcd for
C13H18O3: C, 70.24; H, 8.16%. Found: C, 70.02; H, 8.10%.
3.5. Acetylation of α-hydroxy ester
To a 50 ml round-bottomed flask equipped with a magnetic stirrer, was added 1.5 g (7.2 mmol) (R)-2a,
1.10 g (10.8 mmol) acetic anhydride, and 3–4 ml pyridine. The esterification was run for 5–6 h at room
temperature, as checked by TLC (Rf=0.44, hexane:EtOAc=5:1 as an eluent). After usual work-up and
purification of the product through silica gel column chromatography, the acetylated product 2a-OAc
1
was obtained in 94% yield. H NMR (δ from TMS in CDCl3): 1.26 (t, 3H, J=7.1 Hz), 2.15 (s, 3H),
2.09–2.21 (m, 2H), 2.74 (t, 2H, J=7.3 Hz), 4.19 (q, 2H, J=7.1 Hz), 4.98 (t, 1H, J=6.5 Hz) and 7.13–7.28
(m, 5H).
In a similar procedure, other hydroxy esters were also esterified in 86–95% yields.
2c-OAc: 1H NMR (δ from TMS in CDCl3): 1.26 (t, 3H, J=7.1 Hz), 2.10–2.22 (m, 2H), 2.15 (s, 3H),
2.31 (s, 3H), 2.72 (t, 2H, J=7.2 Hz), 4.19 (q, 2H, J=7.1 Hz), 4.99 (t, 1H, J=6.4 Hz), 6.98 (d, 3H) and 7.15
(m, 1H).
2e-OAc: 1H NMR (δ from TMS in CDCl3): 1.28 (t, 3H, J=7.1 Hz), 2.08–2.19 (m, 2H), 2.14 (s, 3H),
2.82 (t, 2H, J=8.1 Hz), 3.78 (s, 3H), 4.20 (q, 2H, J=7.1 Hz), 4.98 (t, 1H, J=6.3 Hz), 6.82 (d, 2H, J=8.6
Hz) and 7.11 (d, 2H, J=8.6 Hz).
2f-OAc: 1H NMR (δ from TMS in CDCl3): 1.27 (t, 3H, J=7.2 Hz), 2.05–2.18 (m, 2H), 2.16 (s, 3H),
2.70 (t, 2H, J=7.3 Hz), 4.19 (q, 2H, J=7.2 Hz), 4.95 (t, 1H, J=6.4 Hz), 6.95 (m, 2H) and 7.13 (m, 2H).
2m-OAc: 1H NMR (δ from TMS in CDCl3): 1.30 (t, 3H, J=7.1 Hz), 2.10–2.21 (m, 2H), 2.16 (s, 3H),
2.87 (m, 2H), 4.22 (q, 2H, J=7.1 Hz), 5.01 (t, 1H, J=6.3 Hz), 7.38 (d, 2H, J=8.9 Hz) and 8.15 (d, 2H,
J=8.9 Hz).

D. H. Dao et al. / Tetrahedron: Asymmetry 9 (1998) 2725–2737
2735
3.6. Oxidation of phenyl group to carboxylic acid with ruthenium tetroxide
To a 100 ml round-bottomed flask equipped with a magnetic stirrer, was added 2a-OAc (0.5 g, 2.0
mmol), carbon tetrachloride (7 ml), acetonitrile (7 ml), water (10.5 ml) and sodium periodate (3.1 g,
14.5 mmol). The contents of the flask were stirred at room temperature until both phases became clear.
To the flask, either ruthenium(III) chloride or ruthenium(II) oxide (2.2 mol%) was added and stirring
was continued until no aromatic compound could be detected by TLC or HPLC. Normally, the reaction
required about 2–3 h depending on the stirring rate. The reaction mixture was then cooled to 0°C with an
ice-bath, and CH2Cl2 (15 ml) was added with vigorous stirring for 10 min. A deep black color appeared
at this moment. The mixture was then transferred to a 100 ml separating funnel and the organic phase
was separated. The aqueous layer was then extracted with CH2Cl2 (3×15 ml). The combined organic
layers were washed with brine (2×10 ml), dried with anhydrous MgSO4, filtered, and concentrated to
give crude monoethyl α-acetoxy glutarate,33–36 which was then esterified to the corresponding diethyl
ester 3 followed by immediate deacetylation to give 4 as a brownish orange oil in 57–94% chemical yield
before purification by column chromatography on silica gel with hexane:EtOAc=5:1 as an eluent.
1
4: H NMR (δ from TMS in CDCl3): 1.20–1.41 (m, 6H), 1.82–2.02 (m, 1H), 2.09–2.28 (m, 1H),
2.43–2.52 (m, 2H), 2.95–3.06 (bs, 1H), 4.08–4.38 (m, 4H) and 4.21–4.41 (m, 1H). IR (cm−1, CCl4): ν
3495 (broad, O–H), 2939 (C–H), 1738 (C_O) and 1216 (C–O). Anal. calcd for C9H16O5: C, 52.93; H,
7.90%. Found: C, 52.94; H, 7.91%.
For ease of detection on HPLC (Chiralcel OD 25+5), the α-hydroxy group of 3 (50 mg, 0.25 mmol)
was esterified by p-nitrobenzoic acid (ca. 50 mg, 0.29 mmol) in refluxing diethyl ether for 30 min. The
product 4, obtained in 70–90% chemical yield, was subjected to column chromatography on silica gel
1
with hexane:EtOAc=3:1 as an eluent (Rf=0.36). H NMR (δ from TMS in CDCl3): 1.20–1.41 (m, 6H),
1.81–1.99 (m, 1H), 2.00–2.28 (m, 1H), 2.62–2.78 (m, 2H), 4.09–4.40 (m, 5H), 6.96 (m, 2H) and 7.14
(m, 2H). Anal. calcd for C16H19NO8: C, 54.39; H, 5.42; N, 3.96%. Found; C, 54.46; H, 5.45; N, 3.91%.
Retention times of (R)- and (S)-4 were 20.50 and 17.06 min, respectively. Both 3 and 4 were cyclized
to lactone 5 by either standing in pyridine for 2–3 days or gentle refluxing for 1–2 h in ethyl alcohol
(5 ml) with a catalytic amount of p-toluenesulfonic acid monohydrate. Pure lactone 5 was obtained in
76–90% chemical yield after column chromatography on silica gel with hexane:EtOAc=1:1 as an eluent
(Rf=0.50).
1H NMR (δ from TMS in CDCl3): 1.30 (t, 3H, J=7.1 Hz), 2.20–2.44 (m, 1H), 2.48–2.70 (m, 1H),
2.53–2.60 (d, 2H, J=6.3 Hz), 4.26 (q, 2H, J=7.1 Hz) and 4.95 (m, 1H). IR (cm−1, CCl4): ν 3000 (C–H),
1780 (C_O) and 1769 (C_O). Anal. calcd for C7H10O4: C, 53.16; H, 6.37%. Found: C, 53.14; H, 6.45%.
For chlorinated hydroxy esters 2i, j and k, the following procedure was employed. To a 30 ml dried
two-necked round-bottomed flask fitted with a magnetic stirrer were added 50 mg of Pd/C and 0.2 g
NaOH which was dissolved in 2.0 ml ethanol, and the solution was cooled to room temperature. Then,
about 200 mg (0.83 mmol) of either 2i, 2j or 2k in 1.0 ml ethanol was added to the mixture. The flask
was then flushed with hydrogen to displace the inside air and was stirred under a hydrogen atmosphere at
room temperature. The reaction was gauged with TLC until most of the chloro derivative was converted
into 2a.31,32 The product, obtained in 42–57% chemical yield, was then checked by TLC and optical
rotation, and compared to the results obtained from the known (R)-(−)-2a.
3.7. Reduction of α-keto esters to α-hydroxy esters by baker’s yeast
The procedure was essentially the same as described previously.13 To a 25×150 mm glass tube, fitted
with a Teflon® cap, containing 1.0 g of baker’s yeast, 10 ml of diethyl ether or a mixture of diethyl

2736
D. H. Dao et al. / Tetrahedron: Asymmetry 9 (1998) 2725–2737
ether and an appropriate amount of water, was added ethyl 2-oxo-4-phenylbutanoate (1a; 12 mg, 0.058
mmol). The mixture was agitated at 130 rpm at 30°C for 24 h. The organic portions were then collected
by filtration and the yeast was washed with either ethyl acetate or diethyl ether (4×10 ml) and the
extracts were combined with the filtrate. The combined organic portions were concentrated under reduced
pressure. The residue was subjected to Extrelut and silica gel to remove involatile materials prior to gas
chromatographic analysis using an HR-20M column for the measurement of chemical yield, and a CP-
cyclodextrin column for determination of enantiomeric excess.
The chemical yield of the product and the amount of the starting material remaining were evaluated by
comparing the signal area of the product to that of the starting material under the standard conditions. No
other product was detected on VPC/HPLC. Whenever necessary, preincubation and reaction times were
changed appropriately.
It was also found that the addition of an appropriate amount of phenacyl chloride to the preincubation
system is effective at improving the enantiomeric excess of the product. For measuring the effect of the
stereocontroller, the microbe was preincubated in the presence of an appropriate amount of the reagent.
The efficiency of the third reagent was tested under the g/l unit, instead of using the mole unit, for
expressing the amount of the reagent added to the reaction system, because the former unit seems more
practical for synthetic purpose than the latter.
3.8. Medium-scale reduction of some keto esters
To a 500 ml three-necked flask containing 240 ml of diethyl ether and 12 ml of water, fitted with a
mechanical stirrer, a thermometer and a reflux condenser, was added 130 mg phenacyl chloride and 22 g
of baker’s yeast. The mixture was preincubated for either 6 or 10 h at 30°C with an agitation rate of 130
rpm. Then, 250 mg of either 1c, e or f was added and the reduction was allowed to proceed for 24 h. The
organic portion was separated by filtration from the yeast and the yeast was washed with ethyl acetate
(4×60 ml). The washings were combined with the filtrate and the combined solution was concentrated in
vacuo to give an odorous yellow viscous residue. The residue was subsequently washed with water (3×5
ml), then brine (2×5 ml) and dried over MgSO4 before it was subjected to Extrelut® to remove yeast
materials. It was then purified by column chromatography on silica gel with hexane:EtOAc=5:1 or 3:1 as
an eluent. Chemical yields and enantiomeric excesses are listed in Table 3.
Acknowledgements
One of the authors (D.H.D.) thanks the Yokoyama Scholarship Foundation for their financial support.
References
1. Servi, S. Synthesis, 1990, 1.
2. Csuk, R.; Glänzer, B. I. Chem. Rev., 1991, 91, 49.
3. Hanson, R. M. Chem. Rev., 1991, 91, 437.
4. Santaniello, E.; Ferraboschi, P.; Grisenti, P.; Manzocchi, A. Chem. Rev., 1992, 92, 1077.
5. Schmidt, E.; Blaser, H. U.; Fauquex, P. F.; Sedelmeier, G.; Spindler, F. In Microbial Reagents in Organic Synthesis; Servi,
S., Ed. Kluwer Academic Publishers, The Netherlands, 1992; pp. 377–388.
6. Wang, Y.-F.; Chen, S.-T.; Liu, K. K.-C.; Wang, C.-H. Tetrahedron Lett., 1984, 30, 1917.
7. Attwood, M. R.; Francis, R. J.; Hassel, C. H.; Kröhn, A.; Lawton, G.; Natoff, I. L.; Nixon, J. S.; Redshaw, R.; Thomas, W.
W. Fed. Eur. Biochem. Soc. Lett., 1984, 165, 201.

D. H. Dao et al. / Tetrahedron: Asymmetry 9 (1998) 2725–2737
2737
8. Hummel, W.; Schmidt, E.; Wangrey, C.; Kula, M. R. Appl. Microbiol Biotechnol., 1986, 25, 175.
9. Hummel, W.; Schutte, E.; Schmidt, E.; Wandrey, C.; Kula, M. R. Appl. Microbiol Biotechnol., 1987, 26, 409.
10. Chadha, A.; Manohar, M.; Soundararajan, T.; Lokeswari, T. S. Tetrahedron: Asymmetry, 1996, 7, 1571.
11. Blaser, H.-U. Tetrahedron: Asymmetry, 1991, 2, 843.
12. Minamikawa, H.; Hayakawa, S.; Yamada, Y.; Iwasawa, N.; Narasaka, K. Bull. Chem. Soc. Jpn, 1988, 61, 4374.
13. Dao, D. H.; Kawai, Y.; Hida, K.; Hormes, S.; Nakamura, K.; Ohno, A.; Okamura, M.; Akasaka, T. Bull. Chem. Soc. Jpn,
1998, 71, 425.
14. Nakamura, K.; Inoue, K.; Ushio, K.; Oka, S.; Ohno, A. J. Org. Chem., 1988, 53, 2589.
15. Nakamura, K.; Kawai, Y.; Nakajima, N.; Ohno, A. J. Org. Chem., 1991, 56, 4778.
16. Nakamura, K.; Kawai, Y.; Nakajima, N.; Ohno, A. Bull. Chem. Soc. Jpn, 1989, 62, 875.
17. Nakamura, K.; Kawai, Y.; Miyai, T.; Ohno, A. Tetrahedron Lett., 1990, 31, 3631.
18. Deol, B. S.; Ridley, D. D.; Simpson, G. W. Aust. J. Chem., 1976, 29, 2459.
19. Kim, M.-J.; Whitesides, G. M. J. Am. Chem. Soc., 1988, 110, 2959.
20. Kim, M.-J.; Kim, J. Y. J. Chem. Soc., Chem. Commun., 1991, 326.
21. Noller, C. R.; Dinsmore, R. Org. Synth., 1943, Coll. Vol. 2, 476.
22. Harrison, G. C.; Diehl, H. Org. Synth., 1955, Coll. Vol. 3, 370.
23. Smith, L. H. Org. Synth., 1955, Coll. Vol. 3, 793.
24. Miyaura, N.; Ishikawa, M.; Suzuki, A. Tetrahedron Lett., 1992, 33, 2571.
25. Greary, X. J. Org. Chem., 1987, 52, 5026.
26. Middleton, W. J.; Bingham, E. M. J. Org. Chem., 1980, 45, 2883.
27. Corey, E. J.; Schmidt, G. Tetrahedron Lett., 1979, 399.
28. Maurer, B.; Hauser, A. Tetrahedron Lett., 1984, 25, 1061.
29. Price, C. C.; Judge, J. M. Org. Synth., 1973, Coll. Vol. 5, 255.
30. Carpino, L. A.; McAdams III, L. V. Org. Synth., 1988, Coll. Vol. 6, 403.
31. Sone, T.; Umetsu, Y.; Sato, K. Bull. Chem. Soc. Jpn, 1991, 64, 864.
32. Eszenyi, T.; Timar, T. Science, 1990, 20, 3219.
33. Spitzer, U. A.; Lee, D. G. J. Org. Chem., 1974, 39, 2468.
34. Carlsen, P. H.; Katsuki, T.; Martin, V. S.; Sharpless, K. B. J. Org. Chem., 1981, 46, 3936.
35. Nunez, M. T.; Martin, V. S. J. Org. Chem., 1990, 55, 1928.
36. Ferraboshi, P.; Grisenti, P.; Manzochi, A.; Santaniello, E. J. Chem. Soc., Perkin Trans. 1, 1990, 2469.
37. Loewen, P. C.; Makhubu, L. P.; Brown, R. K. Can. J. Chem., 1972, 50, 1502.
38. Manzocchi, A.; Casati, R.; Fiecchi, A.; Santaniello, E. J. Chem. Soc., Perkin Trans. 1, 1987, 2753.